Page 1 of 1

Two HD PVRs... Exercise in frustration

Posted: Tue Jun 15, 2010 10:45 pm
by marc
Well, I'm about to pull my hair out. As I've posted before, I've got two HD PVRs.
Individually they work just fine with minor quibbles.

However, throwing the two together leads to disaster in the same machine.
It starts fine but eventually some live tv will kill off one or both of the units.
Eventually, I'll have to power cycle the hdpvr's or sometimes have to power down the motherboard.

I'll be gleaning the logs to divine what the heck is happening but has anyone had any experience using two of these puppies?

Thanks in advance. I'll also bring this up in the mythtv-users irc room sometime tomorrow.

AUGHHH :cry: H!!!! (Charlie Brown cry)

Re: Two HD PVRs... Exercise in frustration

Posted: Wed Jun 16, 2010 8:07 am
by Patrick
I don't have 2 of these devices but I did have issues when using 2 pci capture cards. Myth would confuse which card was which on random bootups. Maybe you should investigate using udev rules or something similar for usb based devices. With udev you can lock down which system ID each device gets on all bootups. You might want to check the mythtv-users mailing list. I remember reading about this very issue on there a while back.

Re: Two HD PVRs... Exercise in frustration

Posted: Wed Jun 16, 2010 9:07 am
by marc
Nah, it's not a udev rule as the correct devices are pointed to in the mythbackend setup. However you do bring up a chore that I have to address eventually. It would sure beat having to check the devices in the backend after every reboot.

As I said, the devices work just fine at start. One will flake after a new recording starts or a channel change for live TV.

The system, and the mythbackend, remained up so it's not the devices being reassigned device names as they are detected with a boot.

At first, I thought one of the boxes was bad so I tried swapping them. Surprisingly they work just fine when they are the only one on the system. That killed off my simple diagnosis.

I'm hoping it's an issue with firewire channel changing with two cable boxes but I'm thinking that will also be a dead end. But to eliminate this as a cause, I'll switch back to IR control for both boxes to test. (I'll have to make an IR dam for both boxes but thankfully the command IR is multi-channel and I can output signals to each of them without confusing the other.

I'll post my findings as I go along.

The locking of the USB bus is disconcerting when it happens. Perhaps my motherboard is not up to par for this task?